The rates of substance abuse among unemployed black teenagers appear to be substantial and alarming. Etiological paradigms which may help to explain substance abuse among unemployed black youth include 1) life stress, 2) identity/lifestyle, and 3) psychological well-being theoretical paradigms. The proposed study represents one of the first systematic investigations of the etiology of substance abuse among unemployed black youth. Psychosocial predictor measures will be collected at two points in time, and drug use criteria at six points in time, by trained interviewers from 150 black male teenagers who have dropped out of school and are unemployed. Unemployed youth will be solicited through multiple channels, including welfare centers, teen employment and neighborhood youth programs, neighborhood and community leaders, parents attending health and human service agencies, mailings to high school dropouts, and media advertisements. Research solicitations will indicate that unemployed teenagers are being sought for an interview to learn how different youth deal with and experience unemployment. Youth will be informed that some questions focus on substance use. Primary predictor variables to be assessed, based on the three etiological paradigms outlined above, include: 1) negative life events, coping strategies and social support (Life Stress); 2) delinquency, religious commitment, life goals, and purposeful activity (Identity/Lifestyle); and 3) depression, self-esteem, and general self-efficacy (Psychological Well-Being). Validity of standard predictor measures will be assessed from convergence with secondary measures derived from open-ended interview data. Reliability and validity of reported drug use will be examined through 1) consistency of self-report over monthly data collection points, and 2) comparisons of self-report data with collateral reports from peers and significant others. Comparable psychosocial data being collected from current research samples will allow the examination of several sets of secondary hypotheses: Unemployed substance abusers will differ significantly on predictor variables from a) black teen drug abusers referred for residential treatment, b) adjudicated black juvenile offenders, and c) school and church-based normative samples of black teenagers.